<![CDATA[The visual identification of traditional Cirebon batik motifs frequently relies on subjective observation, leading to inconsistent recognition results. To resolve this issue, this study implements a Convolutional Neural Network (CNN) with a four-layer convolutional architecture as an automated classification system. The dataset used in this research contains 1,492 images of Cirebon batik motifs, which are partitioned into a scheme of 80% for training and 20% for validation. Data augmentation is applied during the preprocessing phase to improve the variety and quality of the information processed by the model. The results show that the CNN model achieves an overall accuracy of 92%. Furthermore, the Area Under the Curve (AUC) values ranging from 0.98 to 1.00 confirm the model's strong capability in distinguishing between different motif classes, even though minor challenges persist in identifying motifs with high visual similarities, such as Singa Barong and Paksi Naga Liman. ]]>
Copyrights © 2026
